#emc | Logs for 2005-05-08

[00:00:05] <robin_sz> and it will still be poor accuracy
[00:00:06] <Jymmm> and 100rnd mag's are available =)
[00:00:18] <robin_sz> shrug
[00:00:43] <robin_sz> very useful ... for people who can't hit it with the first shot
[00:01:29] <Jymmm> first you have to define 'it'
[00:01:35] <Jymmm> and is 'it' moving
[00:01:55] <robin_sz> for most of the 10/22 users I have seen, you can define "it" as "barn door"
[00:02:52] <Jymmm> well, 500 rounds for $10 USD isn't too shabby.
[00:03:00] <Jymmm> cheap therapy
[00:03:08] <robin_sz> cheap yeah, and fun I guess
[00:03:17] <robin_sz> let me see,
[00:03:26] <robin_sz> I paid 640 GBP for 5000
[00:03:40] <robin_sz> so �64 for 500
[00:03:48] <robin_sz> 120 USD ?
[00:03:59] <Jymmm> probably
[00:04:42] <robin_sz> but, its the best .22LR ive used
[00:05:27] <robin_sz> holds a 10 shot/13mm group at 50m
[00:07:15] <robin_sz> right .. bedotime
[04:22:16] <les_away> Evening LawrenceG....
[04:23:10] <les_away> Oh well... Bed time..
[04:23:19] <LawrenceG> hello
[04:23:38] <LawrenceG> was just thinking about bed myself
[06:09:38] <A-L-P-H-A> who's around?
[07:10:23] <Phydbleep> * Phydbleep is more of a tall, skinny thing..
[07:57:07] <A-L-P-H-A> Phydbleep still here?
[08:18:09] <asdfqwega> Hehehehehehehehahahahahehehe...
[08:19:45] <asdfqwega> When you ask a girl, "Would you like to see my CNC etchings", and she's impressed...there might be hope.
[08:19:47] <asdfqwega> Maybe.
[08:20:16] <A-L-P-H-A> got a picture of your CNC etchings?
[08:20:23] <A-L-P-H-A> asdfqwega
[08:20:33] <asdfqwega> No, nothing recent
[08:21:29] <asdfqwega> An old friend logged in, and I showed her my old webpage
[08:22:37] <asdfqwega> As an artist who developed tendonitis, the doctor told her NO TYPING
[08:22:47] <A-L-P-H-A> oh
[08:22:51] <A-L-P-H-A> voice recognition.
[08:22:57] <A-L-P-H-A> asdfqwega, got any old CNC etchings?
[08:23:09] <asdfqwega> http://www.bright.net/~janfritz/CNC/My%20DIY%20CNC%20Machine.html
[08:23:54] <asdfqwega> So she's been offline for about a year
[08:24:49] <asdfqwega> I *really* to update that page
[08:24:57] <A-L-P-H-A> you've only made a penguin so far?
[08:25:01] <A-L-P-H-A> oh.
[08:25:35] <asdfqwega> Heck no! I've since stuck a CO2 laser on the thing, and I've been cutting scrollwork in wood.
[08:26:12] <A-L-P-H-A> asdfqwega? You got a co2 laser?
[08:26:14] <A-L-P-H-A> damn.
[08:26:18] <A-L-P-H-A> Where's your CAMERA!?
[08:26:46] <asdfqwega> Uh...I have to clean up in here first.
[08:28:03] <asdfqwega> Which is what I should have been doing...
[08:28:11] <A-L-P-H-A> asdfqwega? where are you again?
[08:28:19] <asdfqwega> Ohio, USA
[08:29:53] <asdfqwega> A-L-P-H-A: where are you?
[08:30:14] <asdfqwega> You're in Toronto, IIRC
[08:30:37] <A-L-P-H-A> Yup, TO.
[08:30:59] <A-L-P-H-A> I should make a little map thingie. Where people can input their locations...
[08:31:48] <asdfqwega> Not only have I been cutting scrollwork and stuff, but I've also been freelancing for a local billiard cue maker who wants custom inlays for his cues
[08:32:29] <asdfqwega> He has a better machine for it, so I CAD the artwork and write the programs
[08:32:48] <A-L-P-H-A> asdfqwega. hmm... I'm having issues with writing CAM things right now.
[08:33:25] <asdfqwega> Yeah, I remember, but I thought you already whipped out MasterCAM and gotten re-aquainted with it
[08:35:32] <A-L-P-H-A> Well, I'm getting the roughing done, but I don't know if it's right. I'm trying to figure out how to makesure it DOES not cut into the workpeice holders.
[08:36:23] <asdfqwega> Are you using a vise? Hopefully of the angle-lock variety?
[08:37:34] <A-L-P-H-A> asdfqwega, no, no vice. I'm gonna clamp it down.
[08:37:41] <asdfqwega> When I was looking over that cover/plug print you showed me, I was hoping that you had a vise with a small step ground into the jaws
[08:38:48] <A-L-P-H-A> http://pastebin.com/281198
[08:38:56] <asdfqwega> a very handy things I've seen and used: A 0.100" by 0.100" area cut out of each jaw, with inner corner relief
[08:39:20] <A-L-P-H-A> asdfqwega, that's cool. I'm making a jig for this... that's not too much of a worry for me.
[08:39:41] <A-L-P-H-A> I'm gonna have to make 50 of them. As there is a part change, i'm gonna go do that.
[08:40:27] <A-L-P-H-A> http://pastebin.com/281198 <-- that roughs out the part... with the top left, and bottom right, are for where I'm going to use clamps.
[08:40:39] <A-L-P-H-A> stock size: 4.25 x 2.5 x 0.5 inches
[08:40:51] <A-L-P-H-A> actually only needs 4"
[08:40:56] <asdfqwega> Well, you could have it rough and finish the top and slots in the vise, and then you have another piece of the same stock that you machine so you can clamp down through the slots, and make it run the outer profile
[08:41:12] <A-L-P-H-A> yeah, that's what I'm thinking.
[08:41:19] <A-L-P-H-A> that's what I'm gonna do i mean.
[08:44:30] <A-L-P-H-A> hi Imperator_. if that's the real Imperator_.
[08:45:45] <Imperator_> of cause
[08:45:51] <Imperator_> morning
[08:46:14] <Imperator_> whats up
[08:48:56] <A-L-P-H-A> what's up?
[08:49:03] <A-L-P-H-A> http://pastebin.com/281198 <-- that roughs out the part... with the top left, and bottom right, are for where I'm going to use clamps.
[08:50:02] <asdfqwega> I'd be better off with a pic of your work area :/
[08:51:05] <asdfqwega> If you don't have the steps in the jaws, a matching pair of parallels
[08:52:05] <Imperator_> can't view the programm at the moment
[08:52:43] <asdfqwega> I need to set up a system where I can use the Axis frontend...preferably a non-RT system
[08:52:54] <A-L-P-H-A> Imperator_, I have parallel bars, not an issue.
[08:53:12] <asdfqwega> And rubber bands
[08:53:53] <asdfqwega> To hold the parallels against the jaws :)
[08:54:25] <asdfqwega> 'course, some people wouldn't want to use rubber bands, oh no...
[08:54:46] <asdfqwega> They would use elastomer retention devices XD
[08:58:15] <asdfqwega> Am I right in assuming that pastebin is a 'online clipboard'?
[08:58:23] <A-L-P-H-A> yes
[08:58:27] <A-L-P-H-A> very handy
[08:58:32] <asdfqwega> Coooool
[09:10:08] <asdfqwega> A-L-P-H-A: Since this is the second time you've mentioned trouble with writing CAM stuff...are you interested in having someone do it for you?
[09:11:47] <A-L-P-H-A> heh.
[09:11:57] <A-L-P-H-A> I'm looking in learning how to do it for myself.
[09:12:25] <A-L-P-H-A> I'd be interested in learning an efficient way to do it automated. So far, everthing that I've written has been by hand.
[09:16:51] <A-L-P-H-A> asdfqwega, what do you use to write gcode?
[09:17:53] <A-L-P-H-A> asdfqwega, got some issues of: MAX rpm my motor can do... "2580rpm" @ 60hz. :(
[09:17:56] <A-L-P-H-A> so sow.
[09:25:28] <A-L-P-H-A> what visual simulator programs are there EMC / bdi?
[09:25:35] <A-L-P-H-A> for emc or on the bdi
[09:27:17] <asdfqwega> There used to be a program emcplot3d, but I don't know that it's been maintained, or even compiled with BDI-4
[09:29:07] <asdfqwega> I'd love to have Axis as either a standalone, or to be able to intall it on a non-RT system, but I don't know that any of the instructions for building a non-RT EMC system will work on the BDI-4
[09:30:52] <asdfqwega> I've even tried installing using the BDI-4 cd as a repository, into a Debian Sarge system - not feasible
[09:32:32] <asdfqwega> Ugh...time for sleep, and possibly food.
[09:33:01] <A-L-P-H-A> food sounds good.
[09:33:05] <A-L-P-H-A> sleep sounds good too
[10:08:07] <A-L-P-H-A> hi Jymmm.
[10:08:08] <A-L-P-H-A> sup?
[10:08:12] <Jymmm> howdy
[10:08:33] <Jymmm> trying to figure out what the official name is for rollerskate bearings
[10:08:45] <Jymmm> AFBMA
[10:09:14] <A-L-P-H-A> ABEC? that's a grade
[10:09:26] <Jymmm> Not ABEC, AFBMA
[10:10:01] <A-L-P-H-A> <shrug>
[10:10:03] <A-L-P-H-A> I just blade.
[10:10:13] <A-L-P-H-A> I could care less what they're called, as long as skate...
[10:25:19] <Jymmm> G'Night, I'll just deal with this inthe morning.
[11:45:27] <Phydbleep> asdfqwega: You alive?
[12:05:33] <Phydbleep> * Phydbleep gives up and goes to fall over..
[12:05:43] <Phydbleep> G'nite everyony. :)
[12:34:43] <anonimasu> morning
[12:59:22] <alex_joni> greetings
[13:01:35] <anonimasu> hey alex
[13:01:41] <anonimasu> how did it work out with kbuild?
[13:01:53] <alex_joni> pretty much ok
[13:02:02] <alex_joni> there is some work needed on the modules themselves
[13:02:11] <alex_joni> but the build process pretty much works now
[13:02:21] <anonimasu> nice
[13:02:21] <anonimasu> :
[13:02:23] <anonimasu> :)
[13:02:30] <alex_joni> the run-scripts need some adjustment too, but that's pretty easy
[13:04:36] <alex_joni> so.. how've you been?
[13:07:05] <anonimasu> I've been ok, been working this morning :)
[13:13:38] <alex_joni> * alex_joni is watching F1
[13:15:31] <anonimasu> nice
[13:15:32] <anonimasu> :)
[13:15:41] <alex_joni> morning les
[13:15:44] <anonimasu> I am amazed how the engines hold togther at that speed
[13:15:53] <alex_joni> yeah.. even with the new rules
[13:16:14] <alex_joni> one engine has to last 2 weekends
[13:16:26] <alex_joni> a set of tires one weekend ;)
[13:16:44] <alex_joni> MSC just changed 2 wheels.. lost a lot of places
[13:20:48] <anonimasu> hehe
[13:24:39] <alex_joni> they can only change the wheels if they are damaged
[13:24:49] <alex_joni> he just had a flat, changed it
[13:24:55] <alex_joni> and a lap later had another flat :(
[13:26:42] <anonimasu> :/
[13:27:08] <alex_joni> seems it's a bad year for ferrari
[13:27:18] <alex_joni> anyways.. back to emc :D
[13:53:20] <dave-e> gosh it's quiet
[13:53:25] <alex_joni> too quiet
[13:53:26] <alex_joni> :D
[13:53:46] <dave-e> like is everyone signed on but asleep
[13:54:47] <alex_joni> yup
[13:54:55] <alex_joni> * alex_joni is fighting with rtapi on 2.6
[13:55:16] <dave-e> oh...well it will keep you out of trouble
[13:55:37] <dave-e> I'm hoping les will show up...
[13:55:45] <dave-e> need to pick his brains
[13:55:48] <alex_joni> I don't get it.. it fails now..worked yesterday :(
[13:56:28] <dave-e> gremlins....like yesterday evening I had a g2 cut as a g3....repeatable
[13:56:36] <dave-e> wrong radius also
[13:56:42] <alex_joni> darn
[13:56:53] <dave-e> suspect it will work ok today.
[13:57:02] <dave-e> think the cpu was too warm
[13:57:26] <dave-e> will find out this morning w/o messing up a good piece of steel
[13:58:06] <dave-e> cradek...you there
[13:58:50] <dave-e> brb .. coffee
[14:00:45] <dave-e> i'm back
[14:01:20] <alex_joni> seems cradek is idle
[14:01:22] <alex_joni> les too
[14:01:30] <alex_joni> dave: you got any Makefile experience?
[14:01:56] <dave-e> ....dave is awful at make...I can read the book but even that doesn't help
[14:02:04] <alex_joni> heh
[14:02:38] <dave-e> do you know of anyone that has tried g2/3 at really high speeds?
[14:02:58] <alex_joni> nope
[14:03:05] <dave-e> thats why I want les
[14:03:21] <dave-e> my thought is: that if it is smooth...
[14:03:58] <dave-e> and I assume it is implemented as a function....and updated every servo cycle...or at least tp cycle..
[14:04:26] <dave-e> then one might get by the jerkiness that les experiences at high speeds..
[14:04:50] <dave-e> by implementing an arbitrary nth-order function as a g code.
[14:05:33] <dave-e> Gnnn coeffecients for the nth order.. end point
[14:05:59] <dave-e> it gets eval on the fly just like circular
[14:07:26] <dave-e> wow! I didn't even get flamed
[14:07:39] <dave-e> guess no one cares
[14:08:12] <alex_joni> no one ere to care
[14:08:21] <alex_joni> * alex_joni is deep in rtapi sht
[14:08:30] <alex_joni> and seems everybody else is not around
[14:08:31] <dave-e> get a shovel
[14:09:05] <alex_joni> hehehe
[14:09:11] <alex_joni> that's how I got in there
[14:09:19] <alex_joni> I need a CAT to dig me out
[14:09:20] <dave-e> this should be prime time....
[14:09:38] <dave-e> oh...bad news...you are in deep
[14:09:59] <dave-e> maybe a 10 yd front loader? ;-)
[14:12:05] <dave-e> lets see alex...you are gmt+2
[14:12:39] <alex_joni> yup
[14:12:47] <dave-e> well I got lucky
[14:12:51] <alex_joni> makes it 17:12 over here
[14:12:52] <alex_joni> :)
[14:13:09] <dave-e> and we're -7 right now
[14:13:45] <dmess> hi all...
[14:13:55] <dave-e> mornin'
[14:14:17] <dmess> has every one made the obligatory call to mom yet...
[14:16:39] <dave-e> Matt and Ray are off working on a lathe
[14:17:12] <dmess> making any progress??
[14:17:26] <dave-e> they had motion a couple of days ago
[14:18:21] <dmess> but is it controlled motion
[14:18:29] <alex_joni> and some IO yesterday afaik
[14:18:47] <dave-e> pretty much..needed to tune
[14:19:06] <dave-e> when I t alked to Matt they were just starting on the i/o
[14:19:24] <dmess> right on.. at least its a lathe.. only 2 axes to tune
[14:19:43] <dmess> is it all emc??
[14:20:01] <dave-e> well, with the right tools tuning usually isn't too difficult
[14:20:18] <dave-e> with Matt or Ray...naturally it is emc
[14:20:46] <dmess> sevro's or steppers??
[14:21:07] <dave-e> I'm sure i t is servos...
[14:21:17] <alex_joni> "Allocation is a fatal error."
[14:21:36] <dave-e> not nice!
[14:21:58] <dave-e> Lathe is a Hardinge so it is worth converting
[14:22:20] <dmess> was it a cnc or a manual hardinge??
[14:23:05] <dave-e> I think it was an old cnc...oh yeh....they replaced the syncros with encoders
[14:25:01] <dmess> superslant 4 axis would be sweet...
[14:25:19] <dave-e> gosh..you don't want much. ;-)
[14:25:29] <dave-e> but yes it would
[14:25:42] <dmess> pie in the sky... in sterio..
[14:25:53] <dave-e> but fun to think about
[14:26:10] <dmess> dream big... and go for it...
[14:26:33] <dmess> every step closer to the dream is an experience
[14:29:27] <dmess> dave-e.. wanna read in intresting High speed manual???
[14:29:41] <dave-e> like?
[14:30:21] <dmess> i just scanned the Toshiba TOSNUC 888 SHAPE manual
[14:30:39] <dave-e> impression?
[14:30:47] <dmess> mine...??
[14:30:55] <dave-e> oh!!!!
[14:31:20] <dmess> its awesome... i was factory for 3 yrs..
[14:31:52] <dave-e> I'll have to see if google produces anything
[14:34:35] <dmess> i can dcc a pdf...
[14:35:29] <dave-e> sure...hit it
[14:37:08] <dmess> 'bout a meg
[14:44:03] <alex_joni> starting to get crowded
[14:44:13] <dave-e> well...finally
[14:44:20] <alex_joni> * alex_joni finally got rtapi.ko to build
[14:44:56] <dmess> alex... wanna read some thing and comment??
[14:45:04] <alex_joni> sure
[14:45:12] <dmess> dcc ok..
[14:45:12] <alex_joni> do I have to think much?
[14:45:14] <dan_falck> dave-e: did you find any way arounde the G43/G49 problem?
[14:45:15] <alex_joni> sure
[14:45:29] <dave-e> yes...just didn't use it...
[14:45:30] <alex_joni> try it.. can't remember if it's blocked or not :)
[14:45:41] <dave-e> paul is going to fix...just not instant
[14:45:49] <dan_falck> ok
[14:46:15] <dmess> whats the g43/g49 issue??
[14:46:43] <dan_falck> it was somehow broken lately-didnt' work at all
[14:46:50] <dave-e> tool length
[14:47:04] <dave-e> made z non modal ... nasty
[14:47:05] <dmess> oh i didnt know we broke it..
[14:47:33] <dmess> ummm... gevel 1 g code... must be modal..
[14:47:42] <dave-e> well it only gets important when you try to use it and the z heads for the bed
[14:48:03] <dmess> dumpster diving...
[14:48:38] <dave-e> estops are nice!
[14:49:13] <dave-e> brb
[14:49:19] <dmess> the friedly mushroom...
[14:50:47] <alex_joni> the nice red mushroom
[14:52:24] <alex_joni> dmess: got it
[14:52:53] <dmess> high speed shape manual from a tosnuc 888 control..
[14:53:02] <alex_joni> nice
[14:53:23] <dmess> we need to implement some of that into emc...
[14:55:08] <dmess> its control program over what a certain part of a tool path is going to be smoothed/or toleranced to
[14:55:09] <nevyn_> nevyn_ is now known as nevyn
[14:55:29] <dave-e> dmess..be nice to know how they implemented that...
[14:56:36] <dmess> its in the manual.. s-type accel...i forget was 3 yrs and 2 crashes ago..
[14:57:47] <dmess> averaging and smoothing..
[14:58:46] <dave-e> lots of parameters to set=control
[14:59:19] <dmess> you could set max feed for the program in the 1st line... and the control would monitor what it was doing @ tool cutting pt and move the machine kinematicly at its best speed..
[15:01:32] <dave-e> gotta go make chips...back later
[15:03:12] <alex_joni> dmess: let's wait till we have a good TP/TC
[15:03:20] <alex_joni> after that we can go on from there
[15:03:40] <alex_joni> I mean a 6-axis cubic (or maybe even quintic) trajectory planner with smothing
[15:04:36] <dmess> this needs to be part of the tp..
[15:04:43] <dmess> not an add on
[15:04:43] <alex_joni> yup
[15:04:46] <alex_joni> I see
[15:04:58] <alex_joni> well.. afaik there is plan to have more than one TP
[15:05:25] <dmess> after thoughts are just patches... this needs to go into the planning of the tp... now..
[15:05:41] <alex_joni> well.. bug les & paul
[15:05:45] <alex_joni> they are working on that
[15:06:06] <dmess> ive made a large enuf pest of myself already... ; )
[15:12:12] <dmess> SF function was cool to use too...
[15:12:19] <les> hi all
[15:12:26] <les> Paul is off to matt's
[15:12:39] <dmess> surface speed control ...
[15:12:50] <les> and to se fred about our tp findings
[15:12:55] <les> see
[15:13:08] <dmess> hello les wanna read and comment on something??
[15:13:15] <les> sure heh
[15:13:20] <dmess> dcc ok??
[15:13:24] <les> yeah
[15:14:02] <les> try that again please
[15:14:50] <les> oops what is the extension?
[15:15:05] <les> I have a blocker blocking it
[15:15:06] <dmess> pdf...
[15:15:14] <les> ok one moment
[15:15:24] <dmess> i can change ext
[15:15:42] <les> no I have it now
[15:15:51] <les> please send again
[15:16:44] <dmess> now you have it
[15:16:55] <les> ok looking
[15:18:00] <les> well has user selectable s-curve prob cubic
[15:18:10] <les> typical for the last ten tears or so
[15:22:52] <les> has the usual angle detector...
[15:23:04] <les> Hi john
[15:24:44] <alex_joni> hey jmk
[15:24:52] <alex_joni> just commited some changes to rtapi.c
[15:24:58] <alex_joni> starting on hal_lib.c now
[15:25:31] <jmkasunich> sorry about the dumb mistakes (location and semicolons)
[15:25:56] <dmess> toshibamachine.ca/press_releases/press12.html
[15:26:03] <jmkasunich> the make system wouldn't even attempt a compile last night so I couldn't check it
[15:26:07] <alex_joni> jmk: no problem ;)
[15:26:20] <alex_joni> fixed the make.. should work now
[15:27:20] <les> dmess: nice. A tiny high speed machine
[15:28:15] <dmess> $$$$$ too
[15:28:21] <les> I'm sure
[15:28:33] <dmess> spindle is 80 000 US
[15:28:53] <dmess> minimum rpm is 8000
[15:29:13] <les> heh
[15:29:49] <dmess> a big cutter on it is 4mm
[15:30:23] <les> yes well it does not have much spindle power
[15:31:07] <les> it's for small stuff
[15:31:31] <dmess> no but it has the axis speeds to keep the chiploads constant.. and warp speed
[15:31:48] <les> hmm 10 meters/min
[15:32:00] <les> that's moving along for metal.
[15:32:49] <dmess> machine is ony 1/2 meter acros... it gets there fast/..
[15:33:02] <les> yup
[15:34:36] <dmess> dmg's linear drive machines will hump over at 3m/s
[15:34:49] <anonimasu> iab
[15:34:54] <dmess> they are stupidly fast..
[15:35:12] <les> I only run 5 M/min due to low spindle hp and emc TP
[15:35:19] <les> I need to run 30 or so
[15:36:58] <les> 30 is typical for the run of the mill wood/plastic cutting machines
[15:37:06] <les> 100+ for the high end ones
[15:37:14] <alex_joni> I wonder why CIA doesn't report commits anymore
[15:37:15] <alex_joni> or is it just me who doesn't see them ?
[15:37:34] <alex_joni> 100M/min ?
[15:37:35] <jmkasunich2> did you just commit?
[15:37:39] <alex_joni> yup
[15:37:41] <alex_joni> hal_lib.c
[15:37:43] <jmkasunich2> I haven't seen any since I started
[15:37:44] <les> I am not seeing any
[15:37:54] <alex_joni> it's since last night
[15:38:02] <dmess> whats a commit??
[15:38:13] <alex_joni> an update to a file in CVS
[15:38:15] <jmkasunich2> putting new code or changes into sourceforge
[15:38:16] <dmess> like Hale bop??
[15:38:24] <les> heh
[15:38:35] <anonimasu> coud you run at 100M/min with emc and a STG/vital/dumb card
[15:38:40] <dmess> oh i see... so wheres the history tree??
[15:38:48] <les> It's also when they give up on you and put you in a rubber room.
[15:38:56] <alex_joni> cvs.sourceforge.net/viewcvs.py/emc
[15:39:09] <alex_joni> an0n: not with this TP
[15:39:18] <anonimasu> dmess: Unrivaled high-precision run-out (0.07 mm)
[15:39:22] <alex_joni> unless you have some very long straight lines
[15:39:39] <anonimasu> that sounds like lots
[15:39:43] <dmess> been there... nice theraputic green walls and w nice whit jacke... with long sleeves..
[15:39:43] <alex_joni> one 100M G1 would be ok
[15:39:59] <anonimasu> that sounds like pretty much runout
[15:40:00] <alex_joni> heh.. and the best thing: you can scratch your own back
[15:40:01] <les> To put things in perspective....emc cannot run at even 1 cm/sec with points .1 mm apart
[15:40:17] <anonimasu> les: with the planner fixed ofcourse
[15:40:22] <les> right
[15:40:27] <alex_joni> you mean with the new planner
[15:40:27] <dmess> why not???
[15:40:28] <alex_joni> :D
[15:40:37] <alex_joni> dmess: it runs out of points
[15:40:48] <alex_joni> jmk: hal should work now
[15:40:48] <les> well paul is on the way back to fred to talk about it
[15:40:53] <alex_joni> * alex_joni is trying an example
[15:41:07] <dmess> as in it looks ahead too much and wons dump its buffer??
[15:41:21] <les> Our little spiral c program makes a great test
[15:41:43] <jmkasunich2> I just committed a fix for the uninitialized variable warnings in hal_motenc.c
[15:43:28] <jmkasunich2> got a load of warnings on hal_lib.c
[15:43:39] <alex_joni> you did?
[15:43:41] <alex_joni> what warnings?
[15:43:57] <rayh> rayh is now known as rayh_lathe
[15:44:05] <jmkasunich2> oh, I see what happened
[15:44:19] <jmkasunich2> hal_lib.c is compiled twice, once for user space, once for realtime
[15:44:33] <jmkasunich2> the EXPORT_SYMBOL calls need to be ifdefed out in user space
[15:44:42] <jmkasunich2> I'll fix it
[15:45:03] <alex_joni> right
[15:45:16] <alex_joni> didn't happen here, because I already had the user space hal_lib
[15:46:21] <jmkasunich2> works better now... still some other warnings
[15:47:14] <alex_joni> yay
[15:47:19] <alex_joni> siggen is siggening
[15:47:20] <alex_joni> :D
[15:47:35] <alex_joni> halmeter doesn't build though :(
[15:47:40] <alex_joni> looking at that now
[15:47:50] <jmkasunich2> ifdef fix committed
[15:48:59] <jmkasunich2> got a couple warnings in stepgen and siggen
[15:49:42] <alex_joni> oopsy.. GTK_VERSION is not set on this BDI
[15:49:43] <alex_joni> :(
[15:50:13] <jmkasunich2> the detection code in ./configure doesn't work?
[15:50:22] <les> Halscope would have been a bit handy for our tests....had to cobble up a quick op amp integrator design to look at velocity profiles off the DAC
[15:50:32] <les> with a scope
[15:50:48] <alex_joni> not sure what GTK is installed
[15:51:13] <jmkasunich2> what BDI are you running?
[15:51:18] <alex_joni> 4.20
[15:51:33] <jmkasunich2> gtk-config --version
[15:51:43] <alex_joni> there is no gtk-config
[15:51:45] <jmkasunich2> I got 1.2.10 here (also bdi-4.20)
[15:51:52] <alex_joni> hmmm
[15:51:54] <alex_joni> really?
[15:52:34] <alex_joni> where is it located?
[15:52:38] <alex_joni> which gtk-config
[15:52:56] <jmkasunich2> just a sec, checking something
[15:52:57] <dan_falck> ray_lathe: how are things going with the Hardinge?
[15:53:16] <jmkasunich2> alex: do dpkg -l | grep gtk
[15:53:35] <jmkasunich2> I bet you don't have libgtk1.2-dev installed
[15:53:46] <alex_joni> libgtk2.0-0
[15:53:50] <alex_joni> libgtk2.0-bin
[15:53:54] <alex_joni> libgtk2.0-comm
[15:53:56] <jmkasunich2> but not the -dev package
[15:54:04] <alex_joni> nope
[15:54:18] <alex_joni> but I installed everything from the BDI
[15:54:22] <jmkasunich2> -dev has the headers and such needed to develop GTK progs, the other libs are just to run them
[15:54:22] <alex_joni> at least I thought so
[15:54:28] <jmkasunich2> did a developers install?
[15:54:29] <alex_joni> yup
[15:54:45] <alex_joni> yes
[15:55:43] <jmkasunich2> I may have explicitly told it to install GTK 1.2, or did an apt-get later
[15:56:28] <alex_joni> apt-get'ing it now
[15:56:45] <alex_joni> we might want to add this to the wiki on BDI4-Install
[15:57:37] <jmkasunich2> yeah
[15:57:47] <alex_joni> jmk: you know what you could use?
[15:57:55] <jmkasunich2> a vacation?
[15:58:01] <alex_joni> some red/green LED's on your compile farm on every slot :)
[15:58:06] <alex_joni> j/k
[15:58:45] <alex_joni> darn sloooow connection
[15:59:06] <jmkasunich2> darn big packages too
[15:59:25] <alex_joni> 1MB is not that big ;)
[15:59:29] <anonimasu> heh
[15:59:48] <alex_joni> http://www.linuxcnc.org/compile_farm/emc2kbuild_slot3.txt
[15:59:55] <alex_joni> we need some fixing of the Makefile
[16:00:49] <jmkasunich2> to declare EXPORT_SYMTAB I think
[16:00:59] <alex_joni> yes
[16:01:43] <jmkasunich2> maybe actually in ./configure, not the makefile
[16:01:58] <jmkasunich2> just add it to CFLAGS (the ones used for realtime compiles)
[16:02:04] <alex_joni> well.. in the Makefile.inc
[16:02:14] <jmkasunich2> yeah
[16:02:19] <alex_joni> ok.. how should it look like?
[16:02:30] <alex_joni> -DEXPORT_SYMTAB=???
[16:02:58] <jmkasunich2> I don't think you need the =, it doesn't need to be defined to anything in particular, just defined
[16:03:07] <jmkasunich2> and not defined for kernel 2.6
[16:03:20] <alex_joni> ok.. can do that
[16:05:31] <anonimasu> hm, emc2 is a joy to compile
[16:06:58] <alex_joni> an0n: nice to hear that
[16:07:09] <jmkasunich2> yeah, makes the work worth it
[16:07:15] <anonimasu> :)
[16:07:20] <jmkasunich2> (alex did most of the work)
[16:07:26] <anonimasu> I grabbed it off cvs and just did ./configure ; make
[16:07:56] <anonimasu> I thought I'd give it another try
[16:08:14] <alex_joni> * alex_joni blushes
[16:08:17] <dmess> major enhancements??
[16:08:17] <alex_joni> that's not true :D
[16:08:19] <anonimasu> I dont know what's wrong with emc1 really..
[16:08:25] <alex_joni> dmess: it's not a mess
[16:08:27] <anonimasu> it starts to lag more and more if I dont restart it..
[16:08:33] <anonimasu> between my parts..
[16:08:40] <jmkasunich2> paul and I set up the initial build system, alex did the autoconf part
[16:08:50] <anonimasu> I ran into a part earlier due to that problem..
[16:08:51] <jmkasunich2> and others pitched in as well
[16:08:53] <anonimasu> ;/Y
[16:09:14] <anonimasu> when I was going to jog 5mm
[16:09:18] <dmess> sounds like a good team project.. ; )
[16:09:35] <anonimasu> I ran it before but I got scared off because of the offset trouble there was
[16:10:02] <anonimasu> the best way to solve that is not to use them ;)
[16:10:25] <dmess> post them out of the program like APT used to do..
[16:10:32] <rayh_lathe> Hi dan_falck.
[16:10:39] <anonimasu> apt?
[16:10:41] <dan_falck> ray_lathe: hi
[16:10:44] <rayh_lathe> Lathe project is progresing
[16:10:52] <dmess> Automatically Programmed Tool
[16:10:55] <anonimasu> ah
[16:10:59] <jmkasunich2> got threading working yet? ;-)
[16:11:04] <rayh_lathe> Matt got the tool encoder running about 2 this morning.
[16:11:07] <anonimasu> dmess: they do work but you have to make sure what mode your machine are in relative/absolute ;)
[16:11:21] <rayh_lathe> Some problems with these devices wanting 12 volt.
[16:11:27] <dmess> old nasa programming language..
[16:11:41] <rayh_lathe> Several zeners and 1, resistors later it goes.
[16:11:53] <jmkasunich2> cool
[16:11:55] <rayh_lathe> Also need 12 volt to run the clippard air valves.
[16:12:06] <dmess> go/to,l1,intof,on,c2
[16:12:17] <jmkasunich2> that probably needs more current than a zener supply is good for
[16:12:38] <dmess> which model Hardinge Les??
[16:12:41] <rayh_lathe> I think jmkasunich is taunting me about threading.
[16:12:54] <jmkasunich2> not taunting, just joking
[16:13:05] <rayh_lathe> HNC
[16:13:31] <rayh_lathe> Paul is do here tomorrow sometime. Will work on 4.20 install.
[16:13:40] <dmess> add an encoder to the spindle and use the rigid tapping cycle ...hint hint
[16:13:52] <rayh_lathe> the univpwm works with yesterday's sf.
[16:13:59] <jmkasunich2> dmess: emc doesn't have rigid tapping yet
[16:14:08] <dmess> hint hint
[16:14:21] <jmkasunich2> ray: did matt get 4.20 installed on that rackmount box?
[16:14:21] <rayh_lathe> I've got two spindle encoders. The old 16 ppr and a new US Digital 1k
[16:14:23] <anonimasu> jmkasunich: "yet" I like how that sounds ;)
[16:14:40] <dmess> you'll need IT to thread on the lathe... same function kids... realize it
[16:14:42] <jmkasunich2> unfortunately it's been "yet" for at least two years
[16:14:55] <rayh_lathe> Um not yet. I've been keeping his nose in the Hardinge control cabnet.
[16:15:13] <dmess> only need the start pulse anyway...
[16:15:28] <jmkasunich2> should only take about 5-10 mins of keyboard time, then about an hour of unattended loading
[16:15:55] <anonimasu> jmkasunich: what should?
[16:15:56] <jmkasunich2> we almost got it done at fest, then somebody who was packing up their stuff unplugged the power strip
[16:15:59] <rayh_lathe> I'll see if I can get him to locate it someplace and do that install.
[16:16:11] <dmess> no way...
[16:16:12] <rayh_lathe> Ouch.
[16:16:12] <anonimasu> :(
[16:16:32] <anonimasu> jmkasunich: but you still know how you were planning on doing it ;)
[16:16:58] <jmkasunich2> wait-a-minnit - I'm not talking about threading, I'm talking about Matt's BDI-4.20 install
[16:17:12] <anonimasu> ah ok
[16:17:24] <dmess> git-r-done... les is gonna need it... what good is a cnc lathe that cant thread??
[16:17:25] <jmkasunich2> I wish I was talking about threading!
[16:17:45] <dmess> ops
[16:18:32] <rayh_lathe> rayh is drooling over the slaving in HAL.
[16:18:40] <dmess> ill write the generic macro... if someone else will work on the i/o 's
[16:18:52] <anonimasu> dmess: macro for threading?
[16:18:58] <anonimasu> hm..
[16:19:19] <dmess> you ever seen the inard of a siemens beast??
[16:19:35] <anonimasu> no
[16:19:40] <alex_joni> jmk: it REALLY works
[16:19:42] <alex_joni> hal that is
[16:19:46] <dmess> everything is a macro... with access to i/o signals..
[16:19:50] <alex_joni> with halmeter connected to siggen ;)
[16:19:56] <jmkasunich2> cool
[16:20:25] <anonimasu> neat
[16:20:37] <jmkasunich2> I still have a couple warnings , siggen and stepgen, when they include math.h
[16:20:39] <anonimasu> hm, I've got emc2 configured now..
[16:20:45] <dmess> a g81 has its own subprogram.. you can chang it to mill out a pocket if you want..
[16:20:48] <alex_joni> lemme check
[16:20:50] <jmkasunich2> and a crapload of warnings on the motmod
[16:20:51] <anonimasu> just need to invert some signals
[16:21:02] <alex_joni> yeah I know about those
[16:21:19] <jmkasunich2> anon: what do you need to invert
[16:21:50] <anonimasu> jmkasunich: the dir pins
[16:21:50] <anonimasu> :)
[16:21:51] <alex_joni> __attribute_used__ redefined?
[16:22:08] <alex_joni> an0n: on the parport?
[16:22:10] <anonimasu> yeah
[16:22:20] <alex_joni> instead of connecting them to pin-2
[16:22:25] <alex_joni> connect it to pin-2-inverted
[16:22:28] <anonimasu> yeah I know..
[16:22:29] <anonimasu> :)
[16:22:29] <alex_joni> or how it's called
[16:22:40] <alex_joni> should be pretty easy :)
[16:23:09] <anonimasu> parport.0.pin-02-out-invert
[16:23:20] <anonimasu> err nah
[16:23:27] <anonimasu> setp parport.0.pin-02-out-invert 1
[16:23:28] <anonimasu> it is ;)
[16:23:32] <alex_joni> I was close
[16:23:40] <anonimasu> yep
[16:24:02] <dmess> close is good in dancing and hand gerenades only... not coding..
[16:24:14] <alex_joni> dmess: this ain't coding
[16:24:18] <anonimasu> maybe I should work on this emc > serial > plc thingie
[16:24:26] <dmess> ok sorry sytax... then
[16:24:27] <alex_joni> do it with emc2
[16:24:32] <alex_joni> heh.. yeah
[16:24:33] <dmess> syntax see
[16:24:44] <alex_joni> but.. it's been a year since I looked at that
[16:24:57] <alex_joni> my memory ain't that good
[16:25:10] <nevyn> anonimasu: doing pulse generation offboard? or something else?
[16:25:14] <dmess> find an old hat and put it on
[16:25:22] <anonimasu> nevyn: running toolchanger/robot/somthing else
[16:25:26] <anonimasu> nevyn: whatever I feel like
[16:25:41] <alex_joni> an0n: what happened to that constant tool force thingie?
[16:26:11] <anonimasu> alex_joni: just have the idea need some way to get feedback in to the pc..
[16:26:14] <dmess> tool load monitoring would be nice..
[16:26:42] <dmess> you need to teah a tool... first ..
[16:27:15] <anonimasu> I teah?
[16:27:21] <dmess> monitor and record drive load.. for optimum tool cuttin conditions
[16:27:21] <anonimasu> err what do you mean?
[16:27:25] <dmess> teach
[16:28:31] <dmess> then you monitor for the load to be in that range... when its not you do a retract and t/c procedure and carry on with its replacement tool
[16:28:48] <nevyn> then you know when you break a tool :)
[16:29:22] <nevyn> dmess: but the problem is you need a setting for timber steel brass aluminium etc.
[16:29:23] <dmess> you will never break a lool with tool monitoring on..
[16:30:26] <anonimasu> * anonimasu nods
[16:30:32] <anonimasu> you could do that with a neural net..
[16:30:38] <anonimasu> but that's probably pretty much overkill
[16:30:40] <dmess> yes... to the parameters being PITA to set up... we used to monitor the first and 2nd off.. and average the #'s... and fun
[16:31:04] <dmess> run
[16:32:18] <dmess> and you have to teach each tool /feature..as one might have different characteritics than the next
[16:33:36] <anonimasu> * anonimasu nods
[16:35:15] <dmess> ok so we define x,y,z boxes or zones in with monitoring params which can be altered from one zone to the next..
[16:35:48] <dmess> at rapid no monitoring
[16:36:03] <anonimasu> define what you mean with zone
[16:36:14] <dmess> need to be part of the plc loop/ladder
[16:37:27] <dmess> 2 pts,(x1,y1,z,) and (x2,y2,z2) form a cube in space... if pos'n is in that box...use these spindle monitoring conditions
[16:37:49] <anonimasu> oh, but do you need to have different loads for different points in space?
[16:38:07] <anonimasu> or are you just talking about mapping the drive current?
[16:38:30] <dmess> you might... if one hole has been cored and one is going in solid in the same part
[16:39:05] <anonimasu> hm, I dont quite get why yu need to do that..
[16:39:10] <dmess> mapping spindle drive current with relation to x-y postion
[16:40:00] <anonimasu> you will have to do that for each part you machine wont you?
[16:40:06] <dmess> for unattended toolchanges prior to making scrap...
[16:40:22] <anonimasu> set up loads for different zones..
[16:40:41] <jmkasunich> alex_joni: looks like defining EXPORT_SYMTAB isn't enough
[16:40:44] <dmess> yes.. ; ) or similar parts/tools use generic settings
[16:40:52] <alex_joni> jmk: what for?
[16:40:58] <anonimasu> hm, I dont understand why you have to do that on a per part basis..
[16:41:02] <jmkasunich> for making it build on 2.4
[16:41:05] <alex_joni> ahhh .. build for < 2.6
[16:41:06] <anonimasu> material basis yes, although not per part..
[16:41:16] <alex_joni> well.. unfortunately I don't have a 2.4 anymore
[16:41:24] <jmkasunich> looking at it now on my 2.4 box
[16:42:00] <dmess> unless your programming is consistent enuf to keep metal removal consistenet from part to part..
[16:42:04] <rayh_lathe> Just copied three pics to www.linuxcnc.org/dropbox/ as H*.jpg
[16:42:38] <anonimasu> dmess: that was the idea with having feedback ;)
[16:43:04] <dmess> spindle monitoring is more a monitoring of the tool cutting action than the machine..
[16:43:20] <alex_joni> rayh_lathe: nice control box
[16:43:34] <dmess> but what to check the feedback too??
[16:43:42] <jmkasunich> power box looks nice, control box suffers from spaghetti syndrome right now
[16:43:50] <jmkasunich> (I imagine it will be neater when finished)
[16:43:57] <dmess> brb
[16:44:03] <rayh_lathe> control box is Matt's <g>
[16:44:30] <anonimasu> dmess: you can do that aswell if you want, although I am more interested in keeping a constant chipload while cutting
[16:44:32] <rayh_lathe> We've got to clean up some when we finish connecting IO.
[16:44:49] <jmkasunich> how bout an overall pic of the machine
[16:45:42] <rayh_lathe> Don't know if I can get that far back but will try.
[16:45:49] <rayh_lathe> brb
[16:45:57] <anonimasu> I just took a pic for having while emc loads..
[16:50:59] <alex_joni> jmk: any ideea what -fno-common means?
[16:51:34] <dmess> anon..dcc???
[16:52:21] <rayh_lathe> Hfront and Hback should be approaching Steve's dropbox now.
[16:53:19] <anonimasu> dmess: thanks
[16:54:11] <dmess> ray is missing this comm
[16:54:47] <dmess> anon... look for the sf section..
[16:55:09] <anonimasu> yeah
[16:55:12] <anonimasu> looking in a bit
[16:55:12] <anonimasu> :)
[16:55:14] <dmess> surface speed - feed mode
[16:55:37] <rayh_lathe> Read This Manual Before Operating is always good advice.
[16:55:48] <dmess> flat spot spindle speeds right up.. steep stuff it slos down
[16:56:21] <dmess> cusps it creeps
[16:56:41] <anonimasu> dmess: nice paper
[16:57:02] <dmess> not mine.. toshiba tosnuc 888
[16:58:17] <dmess> something like this needs to be part of emc if its to be of any commercial use to amyone...
[16:58:36] <dmess> HASS almost give their cnc's away guys..
[16:59:01] <dmess> and they are pc to the bone...
[16:59:25] <dmess> plc is written in C++
[17:01:09] <anonimasu> :)
[17:01:28] <dmess> help any?
[17:01:42] <anonimasu> the manual?
[17:01:51] <dmess> yes
[17:03:05] <anonimasu> a bit
[17:03:53] <alex_joni> jmk: still around?
[17:05:39] <dmess> once you have something monitoring the spindle... things like rigid tapping,threading constant surface feed,orbital boring.. are all capable options with the right machine kinematics
[17:09:41] <rayh_lathe> A bunch of the early shape stuff was done at NIST.
[17:10:02] <rayh_lathe> There are some articles -- Tom Kramer I think.
[17:10:21] <dan_falck> rayh_lathe: Matt has a nice smile for being up all night working on code.
[17:10:31] <rayh_lathe> They were working on using probe to find and orient the program to the block of material to cut.
[17:12:10] <dmess> http://www.lexusplant.com/c_pictures.asp?Table=2&ItemNo=222
[17:12:27] <rayh_lathe> The idea was that you could throw down the raw stock anyplace on the table or indexing axis and cut it.
[17:12:31] <dmess> chk that beast out... i set that up in 2002
[17:13:28] <dmess> sory have a read ray.. i have the macros here too someplace
[17:15:32] <alex_joni> this is cool: http://www.kernelnewbies.org/images/penny.png
[17:16:04] <dmess> all this requires at least a high speed skip signal of input back into the program and plc/motion control module
[17:17:51] <rayh_lathe> I love it. "If the workpiece measuring function is not provided as an option, this screen is not appeared.
[17:18:14] <dmess> janglish....
[17:19:00] <dmess> there is some interpretation to be made of some of these manuals...
[17:19:11] <rayh_lathe> Tom got the basics of that into EMC's interpreter in about 94.
[17:19:36] <dmess> basis in 94....??? where's the rest...?
[17:19:44] <rayh_lathe> His diagrams show rotation as well as translation.
[17:19:44] <alex_joni> back in 95
[17:19:46] <alex_joni> :D
[17:20:18] <rayh_lathe> We would have to write the graphical stuff to get it working.
[17:20:40] <rayh_lathe> The spindle load measurement is really good stuff.
[17:21:14] <rayh_lathe> I experimented with a bit of this analog to feedrate override a while back but never got it to the point where it was reliable.
[17:21:35] <dmess> tosnuc 888 just fyi ... has g54,55,56 to accomodate fanuc programs as written... t888 should use g57 h9xx
[17:22:03] <dmess> with xx being optional parameter turn on to 999
[17:22:43] <dmess> so g57 h901 to g57 h9999 for available fixture offsets
[17:23:30] <rayh_lathe> They do provide a nice operator interface to the probing and setting functions.
[17:24:01] <dmess> it was still quite primitive.. but very functional...
[17:24:14] <dmess> all macro program driven
[17:24:40] <dmess> other than the skip signal needing plc input
[17:24:47] <alex_joni> rayh_lathe: what got discussed about iocontroller at Fest?
[17:24:51] <rayh_lathe> the probe calibration stuff is right out of renishaw.
[17:25:11] <dmess> or marposs.. i have their macros too...
[17:25:25] <dmess> we used both
[17:26:02] <rayh_lathe> alex_joni A group of use worked on classicladder but we didn't get there.
[17:26:09] <dmess> i/o for a radio probe required an xtra input to the plc.. if i recall
[17:27:06] <rayh_lathe> Yes and therein lies the ladder rub.
[17:27:19] <rayh_lathe> We need logic running at motion speed.
[17:27:24] <rayh_lathe> But not all of it.
[17:27:58] <rayh_lathe> When Fred got through drawing logic boxes all over a typical emc, it looked like we needed several
[17:28:04] <rayh_lathe> running at the same time.
[17:28:05] <dmess> ok so there are critical calsses and less critical ones... right
[17:28:16] <rayh_lathe> Exactly.
[17:28:54] <dmess> and eahc process within a class has a priority within it right..
[17:29:10] <rayh_lathe> We didn't know just how to implement what we were drawing.
[17:29:54] <rayh_lathe> I believe that HAL will help a lot if we can set up HALIO right.
[17:29:57] <dmess> needs a top down i/o update of all classes deemed critical...
[17:30:25] <dmess> and its always 1 cycle behind...
[17:30:32] <rayh_lathe> Here is where ladder differs from most pc programing.
[17:30:54] <rayh_lathe> ladder tends to read all signals then logic em then write all outs.
[17:31:16] <rayh_lathe> But if some logics are critical how do you break it up.
[17:31:25] <dmess> that a scan cycle yes..
[17:31:26] <anonimasu> on a real plc you have cycle times for your objects..
[17:31:34] <anonimasu> and cycle types..
[17:31:36] <anonimasu> realtime / cyclic
[17:31:57] <dmess> correct a timer conce hit will hold it up..
[17:32:55] <rayh_lathe> How do classes work in a ladder?
[17:33:21] <dmess> lets hypothetically say everythig is a critical process.... with priorities..
[17:33:28] <rayh_lathe> k
[17:34:31] <rayh_lathe> Would you assign an output to a single priority?
[17:34:38] <dmess> its a ladder... run a cycle tthru ...read... update the registers...(logic)... repeat
[17:35:03] <anonimasu> dmess: depends on what kind of plc's though
[17:35:10] <dmess> its a wrung output but not nececarily a physical one
[17:35:21] <dmess> basic ladder..
[17:35:39] <anonimasu> dmess: yuck
[17:35:46] <dmess> it can be used downstream only
[17:36:01] <anonimasu> dmess
[17:36:21] <anonimasu> dmess: on the plc's I work with you can have write/read data internally to different objects running on the plc..
[17:36:23] <dmess> it may be used as an input to a lower wrung
[17:36:38] <anonimasu> and call functions in other objects..
[17:37:03] <dmess> you writing ladder for this... i doubt it..
[17:37:11] <anonimasu> c/c++-ish
[17:37:24] <dmess> sounds m/s i/o typish
[17:37:30] <anonimasu> what?
[17:38:02] <anonimasu> ladder gets messy when you work with large stuff and have lots of IO and things going on on the same PLC
[17:38:41] <dmess> true.. but in the field sometimes you gotta troubleshoot ladder..
[17:39:56] <rayh_lathe> I was trying to talk FredP and a few others into all machine logic in one place -- ladder in my preference.
[17:40:10] <anonimasu> I dont get what you mean really..
[17:40:14] <anonimasu> rayh_lathe: that'd be nice
[17:40:24] <rayh_lathe> But within the EMC, there is logic up in the canonical areas by the interpreter
[17:40:45] <rayh_lathe> logic in task levels, and logic in rt with limits, homes, and probing.
[17:40:49] <anonimasu> yeah
[17:40:55] <anonimasu> gets messy quick
[17:41:13] <anonimasu> quickly..
[17:41:24] <rayh_lathe> Yes it does. But dmess is right about the maint guy wanting it all in one book.
[17:41:32] <anonimasu> yeah
[17:41:41] <rayh_lathe> I know I do when I approach a new machine.
[17:41:51] <anonimasu> since making a change requires other parts to be changed too..
[17:42:25] <rayh_lathe> Yes. A change to a canonical precondition will require changes in other places.
[17:42:39] <anonimasu> rayh_lathe: perhaps multiple SoftPLC threads.. one RT and a nonRT and some way to pass variables*rant*
[17:42:49] <dmess> read the ladder and look for the boo boo...
[17:42:56] <anonimasu> SoftPLCs(classicladder)
[17:43:00] <anonimasu> or whatever you use..
[17:43:19] <rayh_lathe> I wonder how classicladder handles threading.
[17:43:35] <dmess> sounds like an option ...
[17:43:38] <anonimasu> yeah, I am not nescessarily talking about classicladder..
[17:43:42] <anonimasu> although the PLC approach..
[17:43:43] <rayh_lathe> I can imagine it all being stuffed into HAL but then it does not look like ladder.
[17:44:16] <anonimasu> the stuff that dmess said about macros would work great in a world like that
[17:44:26] <anonimasu> writing macros with I/O to the plc..
[17:44:28] <dmess> doesnt have to look like ladder only work as effectively...
[17:44:50] <anonimasu> * anonimasu dosent really care it can be RUBY ;)
[17:45:00] <anonimasu> yeah, anything that works..
[17:45:06] <anonimasu> ladder has its limitations
[17:45:10] <dmess> yes... it does.. sone Variable are w/r from program.. or plc..
[17:45:10] <rayh_lathe> But the maint guy -- and I'm the worst of em -- needs a familiar tool.
[17:45:31] <dmess> EMC would be the tool..
[17:45:47] <anonimasu> I see loads of good stuff in a approach like that
[17:45:56] <anonimasu> probing/aux functions
[17:46:01] <rayh_lathe> Now FredP said they did graphic reps for the K&T and the guys at GM were happy as hell when the had to find and replace a faulty limit switch.
[17:46:06] <dmess> a ladder is just a buncha IF.. THEn..
[17:46:24] <rayh_lathe> But damn if I want to write all those graphical representations.
[17:46:46] <anonimasu> hehe
[17:46:48] <anonimasu> ;)
[17:46:58] <dmess> someone will have to...
[17:47:02] <anonimasu> but a approach liek that would be hell to code
[17:47:05] <anonimasu> like
[17:47:12] <dmess> you ever found a BRAIN DEAD machine
[17:47:36] <anonimasu> the (RT)SoftPLC would have to be lightning fast
[17:48:00] <rayh_lathe> We could do lightning in HAL for those parts that need lightning.
[17:48:04] <anonimasu> I guess the stuff dmess was talking about could be done with clasicladder..
[17:48:10] <dmess> and had to retype a ladder from paper copy... is one thing... to build it from scratch is almost impossible..
[17:48:12] <anonimasu> maybe
[17:48:32] <anonimasu> can you use classicladder in your Gcode?
[17:48:51] <anonimasu> if ( ladder.aux0 < 5 ) {
[17:48:51] <anonimasu> }
[17:48:57] <dmess> no ladder in the g code
[17:48:58] <rayh_lathe> We've talked about making a NML connector to CL.
[17:49:00] <anonimasu> or well, could that be implemented?
[17:49:17] <anonimasu> that's open up for loads of macro functionality the way dmess was talking about
[17:49:29] <rayh_lathe> It would provide contacts from NML messages in and a coil would send NML out.
[17:49:45] <anonimasu> I like the idea of having macros for most stuff..
[17:49:51] <anonimasu> like probe cycles and things like it
[17:50:15] <rayh_lathe> Would those macros be written in g-code?
[17:50:33] <dmess> me too since i HAVE those macros already written
[17:50:42] <anonimasu> rayh_lathe: g-code + ladder
[17:50:56] <dmess> yes g-code/txt files
[17:51:17] <anonimasu> rayh_lathe: what do you think about that idea?
[17:51:21] <dmess> no ladder anon... just referencing plc variables..
[17:51:26] <rayh_lathe> We could do that now with the M101+ codes.
[17:51:43] <anonimasu> dmess: ladder(hardware logic)
[17:51:51] <dmess> wheres the skip signal coming from??
[17:52:12] <anonimasu> dmess: what signal
[17:52:54] <dmess> say for probing...what tells your drives to stop your touching???
[17:53:08] <anonimasu> the signal from the softplc to your macro..
[17:53:34] <rayh_lathe> probing now is handled by a pin assignment and a motion module that looks at that pin.
[17:53:39] <anonimasu> * anonimasu nods
[17:53:39] <dmess> macro shouldnt stop motion.. plc should.. its faster..
[17:53:43] <anonimasu> yeah
[17:53:59] <anonimasu> well you could have the rt plc just stop the motion..
[17:54:01] <anonimasu> directly
[17:54:09] <anonimasu> with the probe hardware logic in it..
[17:54:09] <dmess> plc feeds back to macro to read current pos'n
[17:54:17] <rayh_lathe> That is the way it's done now all rt.
[17:54:55] <dmess> so all we need are the m101 macros.. right??
[17:55:21] <rayh_lathe> And that is where it gets really messy.
[17:55:37] <dmess> i'll send them to you
[17:55:42] <rayh_lathe> I've not used M101's but you can do most anything with them.
[17:55:48] <anonimasu> hm, I didnt think you could pass anything back to thoose macros..
[17:55:58] <anonimasu> or to them from CL
[17:56:01] <rayh_lathe> even cd / sudo rm -rf *
[17:56:19] <anonimasu> rayh_lathe: I thought you couldnt pass stuff back to emc with them
[17:56:51] <rayh_lathe> I don't really know the limits of their ability
[17:57:02] <rayh_lathe> I need to study before I say much more.
[17:57:07] <anonimasu> that was the limit of the user defined M commands that I got told about
[17:57:48] <dmess> you dont... the macros.. basically record the positions to macro vars... youprocess them and store them in your o'set registers from the macro as well
[17:58:15] <anonimasu> dmess: you cant make them move the machine either
[17:58:32] <anonimasu> dmess: you pass stuff to somthing else.. ./a.out
[17:58:48] <anonimasu> or to a program that generates primes..
[17:59:23] <jmkasunich2> * jmkasunich2 is back
[18:00:08] <anonimasu> :)
[18:00:33] <rayh_lathe> Gotta run and get some lunch.
[18:00:45] <rayh_lathe> * rayh_lathe is away: rayh_lathe
[18:00:55] <anonimasu> I'll be back in a bit aswell
[18:02:05] <dmess> move.. chk.. analyse.. move chk.. analyse...repeat..
[18:02:05] <alex_joni> hey jmk
[18:02:10] <jmkasunich> hi
[18:02:15] <alex_joni> jmk: figured out on those warnings
[18:02:26] <alex_joni> the ones with __attribute_used_
[18:02:26] <anonimasu> wb alex
[18:02:38] <alex_joni> an0n: I wasn't away
[18:02:39] <jmkasunich> what is it?
[18:02:46] <alex_joni> http://cvs.sourceforge.net/viewcvs.py/emc/emc2/src/libnml/posemath/_posemath.c?r1=1.2&r2=
[18:03:26] <anonimasu> jmkasunich: what is?
[18:03:47] <jmkasunich> was talking to alex, about warnings
[18:03:51] <anonimasu> ah
[18:04:13] <alex_joni> seems it's a bug in rtai somewhere, and paul already uses a fix for bdi-4
[18:04:23] <alex_joni> I think we can merge that to head
[18:04:27] <alex_joni> or commit the same fix
[18:04:47] <jmkasunich> I'd move the fix over manually
[18:05:00] <jmkasunich> merging for a single fix like that can get hairy
[18:06:04] <alex_joni> ok.. I'll do it
[18:06:19] <alex_joni> but I still don't have a clue what those Warnings on motmod.ko are
[18:06:19] <jmkasunich> I wonder...
[18:06:46] <jmkasunich> the __attrib_used__ things happens wherever we include math.h
[18:07:04] <alex_joni> yup
[18:07:08] <alex_joni> but math.h is from rtai
[18:07:13] <jmkasunich> we can either fix it wherever we use math.h, or
[18:07:15] <alex_joni> so you can't touch that
[18:07:27] <alex_joni> or make a wrapper
[18:07:29] <alex_joni> but...
[18:07:35] <jmkasunich> we can make rtapi_math.h, that includes math.h, and the fix, and make all modules use rtapi_math.h
[18:07:43] <alex_joni> yup
[18:07:55] <alex_joni> hmmm.. might be smthg
[18:08:07] <jmkasunich> kernel modules are asking for trouble including math.h anyway, that's really a user space header
[18:08:33] <jmkasunich> rtapi_math.h could do something like:
[18:08:41] <jmkasunich> #ifdef ULAPI
[18:08:49] <jmkasunich> #incude <math.h>
[18:08:51] <jmkasunich> #else
[18:09:08] <jmkasunich> /* cruft to handle RT weirdness */
[18:09:15] <jmkasunich> #endif
[18:09:51] <jmkasunich> the "cruft" could be as simple as explicitly declaring _only_ the math functions that we use in realtime
[18:11:56] <alex_joni> strange thing
[18:12:01] <jmkasunich> ?
[18:12:05] <alex_joni> when I try to insmod motmod.ko
[18:12:10] <alex_joni> I get a load of errors
[18:12:28] <alex_joni> but the strangest is: unknown symbol main
[18:12:40] <alex_joni> where the hell does that come from?
[18:12:42] <jmkasunich> not too surprising, when you compiled it you got a load of warnings
[18:12:46] <jmkasunich> dunno about the main thing
[18:13:00] <alex_joni> that sounds like smthg userspace got linked with the rest
[18:13:16] <alex_joni> I tried to make only cubic.ko
[18:13:32] <alex_joni> that compiles, gives the same warnings but loads ok
[18:13:32] <jmkasunich> there's no such thing as cubic.ko
[18:13:38] <alex_joni> there is cubic.c
[18:13:44] <alex_joni> and I made cubic.ko
[18:13:45] <alex_joni> :)
[18:13:53] <alex_joni> and insmod' it
[18:14:11] <jmkasunich> I thought cubic.o was just one of the objs that linked together to make motmod.ko, not a module in itself
[18:14:17] <alex_joni> I know
[18:14:27] <alex_joni> but I used it as a module by itself to see if it works or not
[18:14:32] <jmkasunich> ok
[18:15:01] <jmkasunich> I'm a little disconnected from what you are doing, I've been trying to fix the problem on slot 3
[18:15:07] <alex_joni> doesn't make much sense, since it doesn't contain any MODULE stuff
[18:15:13] <alex_joni> oh
[18:15:24] <alex_joni> any luck?
[18:15:27] <jmkasunich> defining EXPORT_SYMTAB breaks slot 3 (but not defining it breakes it worse)
[18:15:40] <jmkasunich> nope, just getting restarted (I was away for a while)
[18:15:50] <alex_joni> I think we need to define a macro, that doesn't do anything
[18:33:50] <alex_joni> darn
[18:33:54] <alex_joni> I don't get this
[18:35:18] <robin_sz> alex_joni: can you check something for me?
[18:35:31] <alex_joni> if I can
[18:35:45] <robin_sz> ok. I'll /msg you the URL
[18:36:43] <robin_sz> I just need to know that it really is accessible from outside
[18:39:29] <alex_joni> darn.. found it :D
[18:39:32] <alex_joni> * alex_joni is happy
[18:40:09] <jmkasunich> alex_joni: what did you find?
[18:40:25] <alex_joni> what caused those buggersome warnings
[18:40:52] <jmkasunich> "buggersome"? You've been hanging out with Paul for too long ;-)
[18:40:55] <alex_joni> heh
[18:41:01] <alex_joni> you'll never dream of what it is
[18:41:24] <alex_joni> * alex_joni found it pretty hard
[18:41:36] <alex_joni> with a mix between emc2 and emc/bdi-4
[18:41:37] <alex_joni> :D
[18:41:48] <jmkasunich> so what was it?
[18:42:01] <alex_joni> in Makefile.inc there's a rule called LD
[18:42:05] <alex_joni> LD = $(CC)
[18:42:08] <alex_joni> that's wrong
[18:42:27] <alex_joni> I deleted the whole line, so kbuild picks his favorite linker
[18:42:30] <alex_joni> now it works
[18:42:51] <alex_joni> now I need to clean the mess I did
[18:42:59] <alex_joni> I think rm * does that
[18:43:10] <jmkasunich2> sometimes
[18:53:58] <robin_sz> indeed
[18:54:57] <robin_sz> try find . | xargs rm -f
[18:55:08] <alex_joni> heh
[18:55:27] <alex_joni> wb rayh_lathe
[18:55:51] <rayh_lathe> Hey Hi.
[18:56:04] <Phydbleep> * Phydbleep frisbees a Dual P-II mb @ asdfqwega..
[18:56:25] <alex_joni> Fido: still didn't get rid of that?
[18:56:31] <rayh_lathe> rayh_lathe is now known as rayh
[18:57:33] <Phydbleep> He wants it, I just needed time to get the Logan over here and to come up with the $ to ship it to him.
[18:57:44] <alex_joni> jmk: any luck on that EXPORT_SYMBOLS issue?
[18:57:56] <anonimasu> iab
[18:58:01] <jmkasunich> still looking... it's related to MODVERSIONS I think
[18:58:20] <alex_joni> what if we ifdef it out?
[18:58:20] <jmkasunich> works on BDI-Live but not TNG, both 2.4 kernels
[18:58:33] <alex_joni> it doesn't have to do anything on <2.6
[18:58:38] <jmkasunich> let me keep looking for a bit
[18:58:39] <alex_joni> or?
[18:58:46] <alex_joni> ok
[18:58:56] <alex_joni> I'll be busy getting motmod.ko to run
[18:59:02] <alex_joni> or at least trying to
[19:01:45] <robin_sz> ick .. that was a BIG hornet
[19:01:52] <robin_sz> 1.25"
[19:02:04] <jmkasunich> splat!
[19:02:09] <alex_joni> hornet?
[19:02:14] <alex_joni> as in bee?
[19:02:20] <robin_sz> as in wasp, yes
[19:02:30] <alex_joni> eat it
[19:02:40] <jmkasunich> lol
[19:02:40] <robin_sz> I think not
[19:02:42] <alex_joni> lol
[19:06:44] <alex_joni> jmk: still around?
[19:07:03] <alex_joni> does the rtapi.ko build ok for you?
[19:07:04] <dmess> auto tool lenght manual and macros anyone??
[19:07:45] <alex_joni> dmess: for emc?
[19:07:48] <alex_joni> :D
[19:08:04] <anonimasu> dmess: please do ;)
[19:08:23] <dmess> not yet.. but i can dream cant i??
[19:08:36] <jmkasunich> still here
[19:08:53] <alex_joni> never mind.. just did a clean checkout
[19:08:59] <alex_joni> works like it should
[19:09:16] <alex_joni> so.. do you want me to do the rtai_math.h stuff?
[19:09:24] <dmess> i found the faulty cut detection manual too..
[19:09:31] <jmkasunich> sure
[19:09:49] <jmkasunich> I'm learning far more than I wanted to about MODVERSION
[19:09:57] <alex_joni> lol
[19:10:46] <dmess> macro's tobe zipped yet...to those how want to see them..
[19:11:06] <anonimasu> * anonimasu waves
[19:11:16] <anonimasu> that took breakage pdf was great
[19:11:25] <alex_joni> jmk: what's halpr_find_comp_by_name ?
[19:11:33] <dmess> next one is better..
[19:11:40] <anonimasu> hehe
[19:11:42] <anonimasu> :)
[19:12:20] <alex_joni> YAY
[19:12:25] <alex_joni> and TRIPPLE YAY
[19:12:28] <alex_joni> motmod works
[19:12:44] <anonimasu> great
[19:12:49] <alex_joni> at least it gets insmod'ed and complains that servo-thread is 0 nsec
[19:13:16] <anonimasu> *dreams of thoose thread speeds*
[19:14:03] <alex_joni> lol
[19:14:30] <dmess> i have peck tapped using rigid tapping 1-8 12 " deep... ; )
[19:15:48] <anonimasu> ouch
[19:18:47] <alex_joni> jmk: about loadrt
[19:18:56] <anonimasu> 12" is deep
[19:18:57] <anonimasu> :)
[19:19:09] <robin_sz> anonimasu: I have that on VHS!
[19:19:26] <anonimasu> robin_sz: lol
[19:21:02] <jmkasunich> hi alex, I'm back
[19:21:03] <anonimasu> I'd love to see that in real life
[19:21:14] <jmkasunich> (was across the room checking things on the farm
[19:21:21] <alex_joni> yeah.. I got it kinda to the point where it should load
[19:21:27] <alex_joni> but loadrt fails
[19:21:39] <jmkasunich> you mean halcmd loadrt?
[19:21:55] <alex_joni> yeah
[19:21:58] <jmkasunich> not surprized... it looks for the modules in a specific directory, and they're not there
[19:22:02] <alex_joni> I know
[19:22:11] <alex_joni> and probably uses thewrong extension too
[19:23:22] <jmkasunich> that code has always been a hack, and would need modified once we have an "install" target too
[19:23:29] <alex_joni> yeah
[19:23:34] <alex_joni> * alex_joni is back later
[19:23:37] <jmkasunich> starts at halcmd.c:894
[19:23:44] <alex_joni> seen that
[19:23:50] <alex_joni> I'm away for half an hour
[19:23:53] <alex_joni> later
[19:23:55] <jmkasunich> ok
[19:30:28] <anonimasu> dmess: looks great
[19:31:06] <dmess> but is it implementable... into emc
[19:34:23] <anonimasu> I guess they are
[19:34:34] <anonimasu> it's not like it's any black magic..
[19:35:16] <anonimasu> I am looking at the calcs..
[19:35:18] <Phydbleep> * Phydbleep grumbles and puts away the live chickens..
[19:35:57] <dmess> whom else needs to see this stuff...
[19:36:37] <anonimasu> it's a average filter for the sampling
[19:37:14] <anonimasu> and a set of rules defining actions
[19:37:51] <anonimasu> if avg/samples < setting/sampletime
[19:38:06] <dmess> yes...
[19:38:43] <anonimasu> I guess that stuff would be very easy to implement in classicladder..
[19:39:13] <anonimasu> that orbital boring stuff looks very cool :)
[19:39:17] <dmess> is classicladder a module??
[19:39:36] <dmess> orbit is the NEXT generation..
[19:39:44] <anonimasu> next?
[19:40:05] <anonimasu> what's to come on the new machines
[19:40:35] <dmess> ifnwhen the speeds get there i can do it all in orbit... 1 machine does it all solution
[19:40:59] <anonimasu> dmess: I still dont quite get how that works..
[19:41:14] <dmess> that function is only availaby on 1 model of machine in the lineup
[19:41:16] <anonimasu> reading the paper now
[19:41:28] <anonimasu> requires blazing speed dosent it?
[19:41:58] <dmess> synchronised spindle orientation and y-y pos'.
[19:42:13] <dmess> yes...it'll do 24m/min
[19:42:18] <anonimasu> y-y pos?
[19:42:22] <dmess> x-y
[19:42:24] <anonimasu> dmess: I meant for the boring to work..
[19:43:29] <dmess> yup... circular accel to .35G at 24 m/minute limit on boring dia.
[19:44:11] <dmess> she shakes rattles and rolls when you cycle start her for the first time...
[19:44:17] <anonimasu> yeah
[19:44:20] <anonimasu> * anonimasu nods
[19:45:21] <dmess> but its so cool to pull a lathe part out of the horizontal with the keways and bhc's done perfectly in register.. 1 setup
[19:46:05] <anonimasu> lol
[19:46:11] <anonimasu> I hate thoose machines.
[19:46:17] <dmess> well i got a brake job to do
[19:46:30] <anonimasu> they represents all stuff I'll never be able to buy ;)
[19:46:38] <anonimasu> hate-love :D
[19:46:44] <dmess> so we'll make our own..
[19:47:06] <dmess> l8r... ; 0
[19:48:05] <anonimasu> * anonimasu nods
[19:48:17] <asdfqwega> Phydbleep: Whachu want?
[19:48:48] <asdfqwega> * asdfqwega is cleaning shop and other things
[20:06:37] <asdfqwega> Phydbleep: Whachu want?
[20:06:40] <asdfqwega> * asdfqwega is cleaning shop and other things
[20:10:22] <alex_joni> * alex_joni is back
[20:10:26] <Imperator_> Hi Alex
[20:10:31] <alex_joni> hey Martin
[20:12:26] <rayh> Did you get home, Martin?
[20:13:50] <Imperator_> Hi Rayh
[20:14:02] <Imperator_> jep i m back since friday
[20:14:35] <Imperator_> New York was a nice expirience
[20:15:50] <alex_joni> seems emc2 is starting to wobble on his feet
[20:15:50] <alex_joni> :D
[20:16:10] <Imperator_> ??
[20:16:19] <Imperator_> it doesent compile at the moment
[20:16:20] <alex_joni> starting to work on 2.6
[20:16:25] <Imperator_> cool
[20:16:25] <alex_joni> it does ;)
[20:16:37] <Imperator_> you are to fast for me
[20:17:11] <Imperator_> that would be cool, i began to hate redhat 7.2
[20:17:37] <anonimasu> hm..
[20:20:10] <anonimasu> lol
[20:20:58] <alex_joni> what?
[20:22:09] <anonimasu> how can you hate somthing that good
[20:22:14] <alex_joni> lol
[20:22:24] <alex_joni> * alex_joni is getting some weird errors
[20:22:41] <anonimasu> I am helping a friend with samba..
[20:22:48] <anonimasu> but I cant get samba to accept windowsxp accounts..
[20:23:02] <anonimasu> and I cant remember how you get xp to allow you to change what user you connect as
[20:24:13] <alex_joni> an0n: net use -user= (net use help)
[20:24:58] <anonimasu> alex_joni: that's not permanent is it?
[20:25:03] <anonimasu> :(
[20:25:10] <alex_joni> it is
[20:25:26] <anonimasu> until reboot or always
[20:25:31] <alex_joni> but.. you can shift-right click on a program link (and the you have Run-As)
[20:25:42] <alex_joni> select the user you want, and connect with that program
[20:26:12] <anonimasu> hm.. ok
[20:28:49] <Imperator_> anonimasu: what do you want to do exactly ?? setup a domain ?
[20:30:43] <anonimasu> Imperator_: I am trying to get samba to accept my xp accounts..
[20:31:10] <Imperator_> that meany you want to see a shared folder at teh samba box
[20:31:18] <anonimasu> yeah
[20:31:30] <anonimasu> I want windows to ask for the user I want to connect as..
[20:31:34] <anonimasu> or well my friend is..
[20:32:09] <anonimasu> hm, the docs dmess had was very nice reading
[20:32:09] <Imperator_> do you have set up teh user in smbpasswd ?
[20:32:17] <anonimasu> Imperator_: you cant have spaces in usernames..
[20:32:25] <anonimasu> that's the trouble..
[20:32:26] <Imperator_> don't know
[20:32:39] <anonimasu> "Forename Lastname"
[20:32:40] <anonimasu> like that..
[20:32:45] <anonimasu> well I'll look at it tomorrow
[20:32:48] <anonimasu> it's no big deal..
[20:33:05] <Imperator_> try a name without spaces then you will see
[20:33:12] <anonimasu> oh the guy cant change he's username ;)
[20:33:43] <alex_joni> jmk: still around?
[20:33:49] <Imperator_> you need the user in smbpasswd also. not only a account at the linux box
[20:34:11] <anonimasu> yep..
[20:34:13] <anonimasu> that's the trouble
[20:34:24] <anonimasu> you cant have "fasdnkl fasdjl" in usernames ;)
[20:34:30] <anonimasu> spaces and such
[20:34:32] <alex_joni> an0n: if it's a XP Pro, and you want to join a domain you also need a Machine Trust Account
[20:34:32] <Imperator_> smbpasswd -a username
[20:34:39] <anonimasu> Imperator_: I know..
[20:34:59] <anonimasu> I dont want it in the domain
[20:34:59] <alex_joni> * alex_joni gives 10 points to the one who answeres this:
[20:35:05] <anonimasu> I want to access a share on the smb box ;)
[20:35:14] <alex_joni> does motmod use NML ?
[20:35:22] <alex_joni> afaik it shouldn't
[20:35:38] <Imperator_> jep there is a script that you add in the smb.conf, then you can make the account on the windoze box when you join
[20:35:39] <jmkasunich> yeah, I'm still here
[20:35:54] <anonimasu> but I dont want it in the domain ;)
[20:36:00] <alex_joni> jmk: think you're the only one around with some knowledge about emc2 :D
[20:36:08] <anonimasu> alex_joni: no I dont think it is
[20:36:09] <jmkasunich> no, motmod doesn't use nml
[20:36:12] <alex_joni> emc.run does start
[20:36:16] <alex_joni> that's what I thought
[20:36:17] <rayh> what's motmod?
[20:36:20] <Imperator_> the connection betwen the interpolater and motion is not done by nml
[20:36:20] <anonimasu> I looked in the file it dosent have any references to nml
[20:36:23] <alex_joni> MOTION: comes up
[20:36:31] <alex_joni> HAL: comes up
[20:36:49] <jmkasunich> rayh: the emc2 equivalent to steppermod, freqmod, stgmod, and the rest.. emc2 has only one, called motmod
[20:37:01] <alex_joni> HAL creates threads 50000 nsec, 1000000 nsec and 10000000 nsec which is ok
[20:37:19] <alex_joni> but MOTION: tries to set Traj cycle time to 0 nsecs
[20:37:24] <Imperator_> the connection is done by usrmotintf
[20:37:26] <alex_joni> and Servo cycle time to 0 nsecs
[20:37:38] <jmkasunich> hmmmmm
[20:37:46] <jmkasunich> do you have uncommitted changes?
[20:37:55] <alex_joni> and I get weird errors from cms_cfg.cc that it can't open 'emc.nml'
[20:37:58] <alex_joni> jmk: some ;)
[20:38:11] <alex_joni> I tweaked and twisted the runscripts to get it so far
[20:38:11] <jmkasunich> commit em and I'll switch over to my 2.6 box and try it
[20:38:22] <alex_joni> don't think that's such a great ideea
[20:38:30] <alex_joni> it might break a lot
[20:38:48] <jmkasunich> you mean on systems other than 2.6?
[20:38:52] <alex_joni> yup
[20:39:04] <jmkasunich> well you have a choice
[20:39:04] <rayh> Thank jmkasunich.
[20:39:07] <alex_joni> lemme do an cvs diff
[20:39:25] <jmkasunich> 1) do it anyway, work out the 2.6 probs, then go back and fix up for <2.6
[20:39:42] <alex_joni> I copied the modules to rtlib/
[20:39:46] <alex_joni> manually
[20:39:50] <jmkasunich> 2) stop where you are, clean up what you have already done so it works on <2.6, commit that, then move on to the next step
[20:40:05] <rayh> Hi Steve.
[20:40:12] <jmkasunich> hi steve
[20:40:24] <alex_joni> 2. I changed the realtime script to have .ko extension
[20:40:28] <stevestallings> Hi Ray, John, ... looks like the troops have been busy
[20:40:33] <Imperator_> Hi steve, what was the url of your homepage with the stepper stuff
[20:40:34] <alex_joni> hey Steve
[20:40:45] <rayh> Busier than a cat in the litterbox.
[20:40:59] <stevestallings> The commercial stuff is at www.pmdx.com
[20:41:11] <Imperator_> thx
[20:41:14] <anonimasu> hm.. strange
[20:41:17] <anonimasu> emc2 wont start
[20:41:37] <alex_joni> what's it saying?
[20:41:58] <anonimasu> cant open config..
[20:42:09] <alex_joni> hmm.. same here
[20:42:15] <alex_joni> what branch are you using?
[20:42:21] <anonimasu> emc2..
[20:42:24] <anonimasu> the standard one
[20:42:27] <alex_joni> HEAD?
[20:42:37] <anonimasu> can't open ini file /home/an0n/EMC2/emc2/configs/emc.ini
[20:42:37] <anonimasu> yeah
[20:42:40] <anonimasu> I think that's it
[20:42:59] <alex_joni> an0n: I presume 2.4 ?
[20:43:07] <jmkasunich> so is there a file at that location?
[20:43:08] <anonimasu> yeah 2.4
[20:43:10] <anonimasu> yes
[20:43:20] <alex_joni> same error here
[20:43:24] <anonimasu> dosent help when I chmod it..
[20:44:33] <jmkasunich> I might have busted something... hold on while I do an update and look at it
[20:44:37] <anonimasu> ok
[20:46:56] <anonimasu> somthing is
[20:47:04] <anonimasu> I copied a old emc.run and it seems to work
[20:47:30] <jmkasunich> still making here, but I have an idea what it is
[20:47:40] <anonimasu> nope..
[20:47:41] <alex_joni> jmk: care to share?
[20:47:43] <jmkasunich> new new emc.run passes the inifile name to halcmd
[20:47:53] <anonimasu> ah
[20:48:01] <jmkasunich> so that hal can get some info from the ini file
[20:48:03] <anonimasu> can I live wthout it ;)
[20:48:07] <jmkasunich> that might not be working right
[20:48:09] <anonimasu> I removed that part from the HAL file..
[20:48:14] <jmkasunich> make is done, let me try a few things
[20:48:17] <anonimasu> sure
[20:48:35] <anonimasu> hm.. still does the same..
[20:49:02] <anonimasu> um lol..
[20:49:03] <jmkasunich> I have a different problem
[20:49:19] <anonimasu> nope still the same.
[20:49:25] <jmkasunich> I take that back, same problem, followed by different ones
[20:49:27] <jmkasunich> ;-)
[20:49:35] <anonimasu> /home/an0n/EMC2/emc2/rtlib/motmod.o base_period=50000 servo_period=1000000 traj_period=10000000 key=101
[20:49:35] <anonimasu> can't open ini file /home/an0n/EMC2/emc2/configs/emc.ini
[20:49:47] <alex_joni> same here
[20:51:47] <anonimasu> I am curious about "inivar"
[20:52:16] <alex_joni> tkemc seems to fail
[20:52:58] <anonimasu> hm, seems like it's somthing other then the run that fails..
[20:53:11] <anonimasu> some program being called
[20:56:30] <alex_joni> jmk: think I found it
[20:56:34] <alex_joni> it's simio
[20:56:36] <Imperator_> * Imperator_ get's no connection to sourceforge
[20:56:50] <alex_joni> bin/simio -ini configs/emc.ini
[20:56:53] <alex_joni> reports:
[20:57:04] <alex_joni> can't open ini file configs/emc.ini
[20:57:08] <alex_joni> an0n: try this
[20:57:16] <alex_joni> open emc.ini and change simio with io
[20:57:44] <anonimasu> I added some debugging stuff.. into the file..
[20:58:53] <anonimasu> trying now
[20:58:55] <anonimasu> works
[20:59:04] <alex_joni> the io?
[20:59:07] <jmkasunich> you can also pass -v to emc.run for more info
[20:59:14] <alex_joni> -d and -v
[20:59:18] <alex_joni> that's what I use
[20:59:31] <anonimasu> hm..
[20:59:33] <anonimasu> weird
[20:59:37] <alex_joni> what?
[21:00:12] <jmkasunich> can be confusing, because simio is started with &, the run script continues and does other things _before_ simio prints the error message
[21:00:29] <anonimasu> it dosent start
[21:00:40] <alex_joni> yup
[21:00:55] <alex_joni> jmk: seems simio behaves different than io
[21:01:01] <anonimasu> emctask.cc 270: rs274ngc_error: Unable to open file
[21:01:01] <anonimasu> Unable to open file
[21:01:01] <anonimasu> emctaskmain.cc 2553: can't initialize interpreter
[21:01:01] <anonimasu> Shutting down and cleaning up EMC...
[21:05:43] <anonimasu> * anonimasu sighs
[21:05:47] <anonimasu> weird.
[21:06:49] <jmkasunich> I have a hunch it's related to the inifile handling code
[21:06:55] <anonimasu> * anonimasu nods
[21:07:16] <jmkasunich> when I merged Paul's lathefork stuff, it included some changes he made to the inifile parsing stiff
[21:07:30] <jmkasunich> changed it to a c++ class and renamed a few things
[21:07:31] <alex_joni> I think so too
[21:07:44] <jmkasunich> what I don't get is why it didn't cause compile or link time errors
[21:08:02] <anonimasu> alex_joni: have you been playing with classicladder?
[21:08:06] <anonimasu> *curious*
[21:08:27] <alex_joni> an0n: a bit
[21:08:29] <anonimasu> can you pass nml messages from it?
[21:08:41] <alex_joni> nope, that's what paul's been playing with
[21:08:43] <anonimasu> ok
[21:08:44] <anonimasu> :)
[21:08:46] <alex_joni> I connected it to hal
[21:09:28] <anonimasu> well I'll give up for today, I'll have more time to play with emc tomorrow afternoon
[21:09:37] <anonimasu> :)
[21:09:42] <jmkasunich> hopefully I'll find and fix the inifile problem by then
[21:09:48] <anonimasu> yeah :)
[21:09:50] <anonimasu> I am in no hurry..
[21:10:10] <anonimasu> my emc1 still runs ;)
[21:10:13] <jmkasunich> gawd I have C++
[21:10:21] <jmkasunich> gawd I _hate_ C++
[21:10:21] <anonimasu> * anonimasu loves it
[21:10:47] <anonimasu> alex_joni: did you have a look at that tool breakage detector paper dmess had
[21:13:04] <rayh> gotta get some work done here. Catch you later.
[21:23:05] <anonimasu_> re
[21:23:08] <anonimasu_> anonimasu_ is now known as anonimasu
[21:23:58] <alex_joni> jmk: I found it ;)
[21:24:08] <alex_joni> seems you replaced -1 with false
[21:24:15] <jmkasunich> ?
[21:24:17] <jmkasunich> where
[21:24:35] <alex_joni> look at simiocontrol.cc line 124
[21:24:44] <alex_joni> I just put -1 back and it works
[21:24:53] <alex_joni> I think more files are concearned
[21:25:32] <alex_joni> emcsh.cc too
[21:25:47] <jmkasunich> another thing I copied over
[21:25:54] <alex_joni> yup
[21:26:17] <jmkasunich> used to be something like if ( 0 != inifile.open(foo)) {
[21:26:25] <jmkasunich> farg
[21:26:26] <alex_joni> -1 == inifile.opn
[21:26:35] <anonimasu> (!inifile.open(foo))
[21:26:38] <alex_joni> now it's inifile.open == false
[21:26:55] <jmkasunich> == false _or_ == -1 are both dangerous
[21:27:12] <jmkasunich> if success is zero and anything else is failure, that's what should be tested
[21:27:18] <anonimasu> yep
[21:27:25] <alex_joni> emctaskmain.cc is touched too
[21:27:37] <jmkasunich> a bunch of em are touched
[21:27:43] <alex_joni> emcsvr.cc too
[21:27:48] <jmkasunich> blast it... I hate merging other folks stuff over
[21:27:49] <anonimasu> I wish I knew perl..
[21:27:50] <anonimasu> :)
[21:28:14] <alex_joni> while at it.. fix emccanon.cc too (that's still with -1)
[21:28:30] <anonimasu> * anonimasu nods
[21:28:42] <alex_joni> emcsh.cc
[21:28:53] <alex_joni> iosh.cc
[21:29:03] <alex_joni> should I fix those on emc2?
[21:29:17] <alex_joni> problem is I can't really test emc2 here
[21:29:21] <jmkasunich> alex_joni: are you _sure_ that actually fixed it (as opposed to simply ignoring a failed open)?
[21:29:22] <alex_joni> only emc2-kbuild
[21:29:26] <anonimasu> I can test them...
[21:29:34] <alex_joni> lemme try to fix them all.. then we'll see
[21:29:35] <anonimasu> :)
[21:29:40] <jmkasunich> alex_joni: I'll do the fix and test on my 2.4 box before I commit to HEAD
[21:29:56] <jmkasunich> I made the mess, I should fix it
[21:30:02] <dmess> one brake job and test drive outta my face..
[21:30:13] <anonimasu> dmess: dont die on us :)
[21:30:41] <dmess> i put some back in too..
[21:30:57] <alex_joni> rebuilding now
[21:31:19] <dmess> my warrentee is 90 second or 90 feet whichever comes first...
[21:36:04] <alex_joni> nope
[21:36:07] <alex_joni> still has issues
[21:36:08] <alex_joni> :(
[21:36:50] <alex_joni> the first one is emctask.cc 270: rs274ngc_error: Unable to open file
[21:36:53] <robin_sz> meep?
[21:37:05] <anonimasu> same error as my stuff died on..
[21:37:37] <robin_sz> I lurv ebay. nicel little dual P3/700 server for �65 :)
[21:39:20] <jmkasunich> that change to a class for interp obviously busted things
[21:40:07] <Imperator_> alex_joni: you are using BDI4.20 right ??? Do you have the /usr/lib/realtime dir ??? On my BDI-4.18 box i dont have that dir !! Wounders why it is not there ??
[21:41:06] <alex_joni> I have /usr/lib/realtime
[21:41:54] <Imperator_> hm
[21:41:55] <alex_joni> jmk: can you point that class out?
[21:43:00] <alex_joni> that interp.xxx stuff?
[21:43:02] <jmkasunich> src/emc/rs274ngc/rs274ngc.hh:31
[21:47:30] <jmkasunich> what error did you have at lines 270 of emctask.cc?
[21:47:55] <alex_joni> Unable to open file
[21:47:59] <alex_joni> I think it's that == false
[21:48:02] <alex_joni> I missed one
[21:48:14] <alex_joni> at Interp.*ini_load
[21:48:47] <alex_joni> emc2 loaded
[21:48:51] <alex_joni> with tkemc
[21:49:16] <alex_joni> jmk: commiting in a second ;)
[21:49:19] <robin_sz> congratulations
[21:49:43] <alex_joni> well.. it just crashed on me.. but hey
[21:49:45] <alex_joni> it's a start
[21:50:00] <jmkasunich> you're on 2.6, right?
[21:50:07] <alex_joni> yup
[21:50:19] <jmkasunich> I though we were talking about the inifile problem on 2.4
[21:50:25] <anonimasu> we are
[21:50:25] <alex_joni> it's the same
[21:50:25] <anonimasu> :)
[21:50:36] <alex_joni> because the branch uses the stuff from head
[21:50:40] <alex_joni> if it's not redefined
[21:50:45] <jmkasunich> right
[21:50:58] <jmkasunich> so you are committing fixes for the branch?
[21:51:00] <alex_joni> so I first thought it's smthg to do with 2.6
[21:51:02] <alex_joni> nope
[21:51:06] <alex_joni> commiting to HEAD
[21:51:09] <jmkasunich> ok
[21:52:29] <anonimasu> hm
[21:52:29] <anonimasu> ok
[21:52:41] <anonimasu> tell me when I can grab it off cvs
[21:53:11] <alex_joni> working as fast as I can
[21:53:29] <anonimasu> np..
[21:53:29] <anonimasu> :)
[21:53:43] <anonimasu> I am not in a hurry
[21:53:53] <anonimasu> just slacking and waiting for bedtime
[21:57:28] <alex_joni> that should do it... I started commiting
[21:58:46] <alex_joni> will probably take a while for cvs to check them out, merge and resend the diffs
[21:58:59] <jmkasunich> huh?
[21:59:05] <alex_joni> darn connection.. I really need a faster one
[21:59:21] <CIA-4> 03alex_joni * 10emc2/src/emc/ (8 files in 4 dirs): changed the test inifile.open() == false to inifile.open() != 0 as it seems that false is not always recognized as it should
[21:59:22] <alex_joni> jmk: on a dial_up here
[21:59:28] <jmkasunich> what do you mean by merge tho?
[21:59:34] <alex_joni> diff
[21:59:36] <alex_joni> not merge
[22:00:50] <alex_joni> an0n: safe to checkout
[22:00:58] <anonimasu> ok
[22:01:00] <alex_joni> hmm.. do you have dev-access?
[22:01:16] <alex_joni> or using pserver?
[22:01:35] <anonimasu> nope..
[22:01:36] <anonimasu> neither
[22:01:48] <anonimasu> I just checkout from sourceforge..
[22:01:49] <alex_joni> how do you access cvs.sourceforge.net ?
[22:02:02] <anonimasu> the public one
[22:02:03] <robin_sz> over ssh usually
[22:02:08] <robin_sz> odh, pserver
[22:02:08] <alex_joni> robin: nope
[22:02:12] <alex_joni> the pserver one
[22:02:14] <anonimasu> ah ok
[22:02:17] <alex_joni> an0n: bad news
[22:02:24] <anonimasu> yeah it's psever..
[22:02:28] <alex_joni> it'll take a while for you to see the changes
[22:02:30] <jmkasunich> it will be a few hours
[22:02:38] <alex_joni> that's why I asked
[22:02:54] <alex_joni> developers get the changes instantly, but public lags a few hours behind
[22:02:59] <anonimasu> yeah
[22:03:45] <alex_joni> you might wanna do it yourself if you're not that pacient :)
[22:04:03] <Phydbleep> * Phydbleep asks that everyone remind him NEVER to move a SouthBend 11x60 again...
[22:04:05] <jmkasunich> alex_joni: looks like it works on 2.4
[22:04:33] <jmkasunich> I know some cvs-foo that will let me merge those changes you just did in head over to the kbuild branch
[22:04:54] <jmkasunich> without disrupting the other stuff we've already done there
[22:05:00] <alex_joni> the best thing is .. it works on 2.6 too ;)
[22:05:02] <anonimasu> I am patient enough..
[22:05:06] <anonimasu> I can grab a new one tomorrow morning
[22:05:17] <alex_joni> well.. do it ;)
[22:05:36] <jmkasunich> right now?
[22:05:36] <alex_joni> seems 2.6 has some issues
[22:05:54] <alex_joni> nah... I'm going to sleep
[22:05:58] <alex_joni> it's 1 am again ;)
[22:06:05] <jmkasunich> (my wife is patiently waiting for dinner.. I'll do it later this evening)
[22:06:07] <alex_joni> I started tkemc
[22:06:18] <alex_joni> and it works, could even home axes
[22:06:26] <jmkasunich> just do a cvs up on your kbuild checkout tomorrow
[22:06:29] <alex_joni> but it seems that it freezes when I try to move
[22:06:34] <Imperator_> les: can you ask Paul why on my BDI 4.18 box is no /usr/lib/realtime directory ?? I have made a standart install
[22:06:40] <alex_joni> ahhhh.. I forgot
[22:06:49] <alex_joni> I removed all the HAL stuff from the emc.ini
[22:06:52] <alex_joni> might be a reason ;)
[22:06:58] <jmkasunich> jmkasunich is now known as jmk_dinner
[22:07:08] <alex_joni> well.. gnight folks
[22:07:37] <alex_joni> was a nice day today
[22:07:47] <Phydbleep> G'nite alex_joni :)
[22:07:49] <anonimasu> yeah it was
[22:07:50] <anonimasu> night alex
[22:07:51] <anonimasu> :)
[22:08:33] <alex_joni> does cia work for you?
[22:08:33] <alex_joni> http://cia.navi.cx/
[22:09:46] <Phydbleep> Bad Gateway
[22:09:46] <Phydbleep> The proxy server received an invalid response from an upstream server.
[22:09:46] <Phydbleep> Apache/2.0.52 (Gentoo/Linux) mod_ssl/2.0.52 OpenSSL/0.9.7d Server at cia.navi.cx Port 80
[22:09:50] <alex_joni> yup
[22:09:53] <alex_joni> same here
[22:09:56] <alex_joni> anyways..
[22:10:00] <alex_joni> good night
[22:10:03] <alex_joni> catch you later